As the book progresses the style becomes more fixed, until in the third decade it has reached its highest point; in the later books, as we know from testimony as well as the few specimens that are extant, it had become garrulous, like that of an old man. His work was to have consisted of fifteen decades, but as we have no epitome beyond Book CXLII., it was probably never finished. Perhaps the loss of the last part is not so serious as it seems. We have thirty books complete and the greater part of five others; but no more, except a fragment of the ninety-first book, has been discovered for several centuries, and in all probability the remainder is for ever lost. Livy was so much abridged and epitomized that during the Middle Ages he was scarcely read in any other form. Compilers like Florus, Orosius, Eutropius, &c. entirely supplied his place. A word should perhaps be said about POMPEIUS TROGUS, who about Livy's time wrote a universal history in forty-four books. It was called _Historiae Philippicae_, and was apparently arranged according to nations; it began with Ninus, the Nimrod of classical legend, and was brought down to about 9 A.D. We know the work from the epitomes of the books and from Justin's abridgment, which is similar to that of Florus on Livy. Who Justin was, and where he lived, are not clearly ascertained. He is thought to have been a philosopher, but if so, he was anything but a talented one; most scholars place his _floruit_ under the Antonines. He seems to have been a faithful abbreviator, at least as far as this, that he has added nothing of his own. Hence we may form a conception, however imperfect, of the value of Trogus's labours. Trogus was a scientific man, and seems to have desired the fame of a _polymath_. In natural science he was a good authority, [55] but though his history must have embodied immensely extended researches, it never succeeded in becoming authoritative.